Promovierender

Tilman Beck M.Sc.

FB 20: Informatik

Ubiquitous Knowledge Processing (UKP) Lab

Kontakt

work +49 6151 16-57575

Work S4|24 108
Dolivostraße 15
64293 Darmstadt

Research interests

  • Natural Language Processing (NLProc) with focus on
    • argumentation mining
    • unsupervised text clustering
    • text summarization
    • deep learning in NLProc
  • Besides that I have a natural interest in
    • (automatic) language learning
    • ethics in NLProc

PhD project

Aggregation and temporal analysis of public discourse about critical infrastructures (working title)


Poster


Given the growth of online mass communication (e.g. social media) in modern society, we can observe public discourse about the functionality of critical infrastructures.

Especially the occurrence of outages or restricted uses of those infrastructures can be observed almost in real-time in this kind of data. With the recent advancements in natural language processing through the use of deep learning techniques, we can develop models to foster the understanding of human-generated text and analyze this data on a large-scale.

The focus of my PhD project is the development and application of computational models for the detection and analysis of public discourse. First, given the vast amount of available data, I want to explore how to aggregate this information into redundant-free and meaningful summaries. Further, I am interested in analyzing the temporal evolution of public discourse on critical infrastructures.

Since 2019 Doctoral Candidate UKP, Technische Universität Darmstadt
2015 – 2018 M.Sc. Computer Science, Christian-Albrechts Universität Kiel, Germany, 2018
2014 – 2015 M.Sc. Betriebliche Umweltinformatik, HTW Berlin, Germany (wish to change back to university)
2013 ERASMUS Semester, Budapest University of Technology and Economics, Hungary
2010 – 2014 B.Sc. Computer Science, Friedrich-Alexander Universität Erlangen, Germany
09/2018 – 04/2019 Research Internship, Technische Universität, Darmstadt
  • argumentation mining
  • unsupervised machine learning
  • deep learning
06/2016 – 01/2018 Student Assistant, Christian-Albrechts Universität Kiel, Kiel 
  • assistance in the research group “Knowledge Discovery”
  • text mining, document clustering, information retrieval
04/2015 – 06/2015 Student Assistant, HTW, Berlin
  • Development of a crowdsourcing-based web application for notification and monitoring of water levels in flood areas
  • Android, Java, JS, data visualization
05/2012 – 09/2014 Student Trainee Software Development, LA2 GmbH, Erlangen
  • Development of individual large-scale web applications for process automation
  • JSP, JS, HTML5, CSS3

Summer Term 2019: Ethics in Natural Language Processing (Tutorial)

Classification and Clustering of Arguments with Contextualized Word Embeddings

Nils Reimers, Benjamin Schiller, Tilman Beck, Johannes Daxenberger, Christian Stab, Iryna Gurevych:
In: The 57th Annual Meeting of the Association for Computational Linguistics (ACL 2019), Florence, Italy, 28.07.2019-02.08.2019, 2019


Survey and empirical comparison of different approaches for text extraction from scholarly figures

Falk Böschen, Tilman Beck, Ansgar Scherp
In: Multimedia Tools and Applications, 2018


Performance Comparison of Ad-Hoc Retrieval Models over Full-Text vs. Titles of Documents.

Ahmed Saleh, Tilman Beck, Lukas Galke, Ansgar Scherp
In: 20th International Conference on Asia-Pacific Digital Libraries, ICADL 2018, Hamilton, New Zealand, November 19-22, 2018, Proceedings, 2018


What to Read Next? Challenges and Preliminary Results in Selecting Representative Documents

Tilman Beck, Falk Böschen, Ansgar Scherp
In: Database and Expert Systems Applications – DEXA 2018 International Workshops, BDMICS, BIOKDD, and TIR, Regensburg, Germany, September 3-6, 2018, Proceedings, 2018

Hack The News Datathon 21-29 January 2019: winning team by UKP, TU Darmstadt

2020 – 2022 Participant in the Software Campus program (funded by BMBF)